Last week I attended the Distributor Conference for RES Software, one of the multitude of Citrix/VMWare eco-system vendors. If you haven't heard about RES, you should really take a look - it's one of the software makers that doesn't get half the respect it deserves. And I don't say that just because COMPUTERLINKS works with them as a distributor, but because I know what a great product they offer.
RES are a Dutch company based in the south of the Netherlands, not too far from Eindhoven. They actually have an enormous feature list but only (currently) two products: Powerfuse and Wisdom. Oddly, the huge list of capabilities represents as much of a challenge as it does an opportunity. Whilst a particular feature may prove a deal-clincher for one company, it may be totally irrelevant for the next. On the other hand, with the software capable of solving so many problems, it is highly unlikely there won't be one killer feature in there for everyone, thus pretty much any company that uses IT is a potential target.
In the same way as Citrix XenDesktop releases the operating system from its dependence on the device (as XenApp does with applications), Powerfuse essentially gives the "user workspace" independence from the operating system. By a workspace, RES mean all the personal settings that a user has, everything from the background picture of their kids to mapped drives and printers and email signatures. This is a very powerful ideology, particularly with the advent of virtual desktops, or VDI, and, as such, it is no surprise that RES count VMWare just about as important a partner as Citrix. Having said that, RES doesn't actually care how you deliver/deploy/install your apps, it works for every way.
Personally, I believe Citrix should have bought RES rather than the Sepago technology they did acquire to solve the problem of roaming profiles, they would have got an awful lot more functionality in RES. Brian Madden counts user workspace management as one of the 5 things he thinks need to happen before VDI can really take off. I have no idea why they didn't buy them, but then I am not privy to that sort of information. I can only presume they would've been too expensive, I would be surprised if the reason had had anything to do with the actual product.
The 2nd product, Wisdom, is a runbook automation product that relieves administrators of many painful and laborious tasks, such as setting up new user accounts, automating provisioning of resources, turning off devices overnight etc. Some of the techies here at COMPUTERLINKS worked out you could save the cost of Wisdom licenses in 6-9 months, just by turning off a few PCs at night and saving the energy they sap. Software that pays for itself in 9 months on one single feature is pretty good in anyone's book.
To be honest, some of what RES offers can be done in other ways, for example with group policies and Active Directory, however the simplicity of RES, added to the sheer amount of functionalities available, are certainly the major buying factors. Security is another one. A good example might be giving a user (doctor perhaps) a specific USB key (by registering the serial number of the USB to that person) and only allowing access to the system when the key is plugged into the device. As I say: possible with other products of course, but there really aren't any other vendors out there who can do so much for so little money.
RES is not without its challenges, not least in terms of awareness. Many have simply not heard of them. AppSense is RES' main competitor, particularly in their home market of the UK, however RES do offer one big advantage for British resellers and distributors and that is that, somewhat refreshingly, they refuse to take end user deals direct. AppSense are famous, or perhaps infamous, within the channel for betraying their partners when they feel like it.
I can't do RES any real justice on a short blog, except to recommend a trial. See for yourself. As regular readers will know, my blogs are fiercely independent and I wouldn't laud a product so openly if I didn't genuinely believe it myself.
Wednesday, 26 November 2008
Wednesday, 5 November 2008
It's a NetScaler Jim, but not as we know it
I think it's fair to say the NetScaler product line has never really been regarded by most as an integral part of the Citrix portfolio. Or at least the technology behind it is not typically something you would associate with Citrix. The buyers are different, the established Citrix channel partners don't always have the networking background required for it and Citrix's arch-rivals in this space, F5, do not compete on anything else that Citrix do. Citrix bought NetScaler a few years ago as a result of the great name it had made for itself powering most of the top portals on the web (Google, Amazon, MSN, Betfair and so on). Citrix rarely hesitate to remind you that 75% of Internet users go through a NetScaler box on any given day. However, the considerable success in the dot-com arena has not been replicated on the corporate side and F5 have had the finance sector, in particular, pretty much sewn up. Until recently that is. Such things as Microsoft Sharepoint and other vendors basing their future innovations very much around web-based front ends are pushing NetScaler's talents further into the limelight. NetScaler revenues from corporates are, for the first time ever, now superceding the dot-com results. Citrix's latest initiative is to build on recent customer successes in accelerating XenApp farms and, by convincing some of the more archetypal Citrix partners to take it on, try and fold this hitherto apparently undesirable stepchild into the Citrix family bosom.
The approach is to be two-fold. NetScaler is essentially a load balancer but has a huge array of more advanced capabilities such as offloading SSL traffic, application firewalling (from the purchase of Teros), multiplexing TCP connections, SSL VPN, caching and compression, to mention but a few. Often, a customer will buy NetScaler for just one or two of these capabilities but end up using more and more of them once the box is in and they become familiar with it.
XenApp is a domineering software, both in the sense that it is resource hungry as well as being a mission-critical, infrastructure-level software so load balancing is critical to performance. XenApp has a much more refined system of assigning users logging in to servers than Microsoft's default "round-robin" system, however it becomes problematic when you need to take things further and introduce fail-over data centres or more advanced DR capabilities for example. Global server load balancing and pro-active, on-the-fly content switching play a large part in enabling a more fault-tolerant, disaster-proof infrastructure, not to mention performance enhancements for the users. NetScaler can also improve XenApp farm performance considerably by relieving the servers of a lot of the XML brokering required to render applications and, of course, accelerating Web Interface, which a lot of companies now use instead of Program Neighborhood (sic).
This is, perhaps strangely, a relatively untapped market for Citrix but one with lots of potential and, crucially, one in which F5 are always going to be on the back foot. Citrix have optimised NetScaler for use with XenApp and competitors naturally aren't going to be able to compete. It is common knowledge that Citrix have over 200,000 customers. For non-disclosure reasons, I can't say how many of those currently employ NetScaler to do what I have outlined in this article, but we were told the figure at Summit last week and it is low.
The second factor that should help align NetScaler more closely with the general Citrix message is the dynamic data centre scenario that has been talked about for a while now. I wrote an article a few months ago about "green IT" (which I published on this blog in September) where I talked about how, in a perfect world, the use of servers (or even whole data centres) could be, or perhaps even should be, aligned more precisely to actual consumption requirements. Basically, I was advocating turning on boxes when they're needed and switching the things off when they're not. NetScaler could play a vital role in this.
The GSLB option does what you'd expect. In the event of a data centre fall-out, all traffic is re-routed elsewhere with no need for any DNS changes. Not only that, it also provides pro-active optimisation of traffic so users get the best service possible, regardless of where they are in the world. Add into the mix the other things NetScaler excels at - policy-based access control, SSL VPN, app firewalling (PCI DSS anyone?) etc. - and NetScaler arguably becomes not just integral, but actually the lynchpin of the truly dynamic data centre. Which is something of an improvement on being an unwanted stepchild.
And finally, we must remember the latest IT buzzword on everyone's lips: "cloud computing". Citrix announced at VMWorld that they now not only have the Citrix Delivery Center (sic) but also C3 - Citrix Cloud Center. This is an amalgamation of XenServer, WANScaler, Workflow Studio and, of course, NetScaler, designed specifically for this new wave of computing that many expect to overtake current architectures relatively quickly. (But cloud computing is a whole topic in its own right and, as such, shall be reserved for a future blog where I hope to also include some reflections on the recent tie-up with Akamai as I believe this could be crucial.)
Will all this help to sell more NetScalers? Almost certainly. I've not even talked about:
1) the fact that Citrix have developed configuration wizards to ease NetScaler deployments,
2) how easy it can be to convince an Access Gateway customer to upgrade to a NetScaler,
3) the imminent release of v9.0 where we can expect specific application integration and
4) that training programs are going online and will cost about 20% of what they used to.
There's a whole lot more to NetScaler than initially meets the eye, it's just a question of whether resellers (and customers) will see it.
The approach is to be two-fold. NetScaler is essentially a load balancer but has a huge array of more advanced capabilities such as offloading SSL traffic, application firewalling (from the purchase of Teros), multiplexing TCP connections, SSL VPN, caching and compression, to mention but a few. Often, a customer will buy NetScaler for just one or two of these capabilities but end up using more and more of them once the box is in and they become familiar with it.
XenApp is a domineering software, both in the sense that it is resource hungry as well as being a mission-critical, infrastructure-level software so load balancing is critical to performance. XenApp has a much more refined system of assigning users logging in to servers than Microsoft's default "round-robin" system, however it becomes problematic when you need to take things further and introduce fail-over data centres or more advanced DR capabilities for example. Global server load balancing and pro-active, on-the-fly content switching play a large part in enabling a more fault-tolerant, disaster-proof infrastructure, not to mention performance enhancements for the users. NetScaler can also improve XenApp farm performance considerably by relieving the servers of a lot of the XML brokering required to render applications and, of course, accelerating Web Interface, which a lot of companies now use instead of Program Neighborhood (sic).
This is, perhaps strangely, a relatively untapped market for Citrix but one with lots of potential and, crucially, one in which F5 are always going to be on the back foot. Citrix have optimised NetScaler for use with XenApp and competitors naturally aren't going to be able to compete. It is common knowledge that Citrix have over 200,000 customers. For non-disclosure reasons, I can't say how many of those currently employ NetScaler to do what I have outlined in this article, but we were told the figure at Summit last week and it is low.
The second factor that should help align NetScaler more closely with the general Citrix message is the dynamic data centre scenario that has been talked about for a while now. I wrote an article a few months ago about "green IT" (which I published on this blog in September) where I talked about how, in a perfect world, the use of servers (or even whole data centres) could be, or perhaps even should be, aligned more precisely to actual consumption requirements. Basically, I was advocating turning on boxes when they're needed and switching the things off when they're not. NetScaler could play a vital role in this.
The GSLB option does what you'd expect. In the event of a data centre fall-out, all traffic is re-routed elsewhere with no need for any DNS changes. Not only that, it also provides pro-active optimisation of traffic so users get the best service possible, regardless of where they are in the world. Add into the mix the other things NetScaler excels at - policy-based access control, SSL VPN, app firewalling (PCI DSS anyone?) etc. - and NetScaler arguably becomes not just integral, but actually the lynchpin of the truly dynamic data centre. Which is something of an improvement on being an unwanted stepchild.
And finally, we must remember the latest IT buzzword on everyone's lips: "cloud computing". Citrix announced at VMWorld that they now not only have the Citrix Delivery Center (sic) but also C3 - Citrix Cloud Center. This is an amalgamation of XenServer, WANScaler, Workflow Studio and, of course, NetScaler, designed specifically for this new wave of computing that many expect to overtake current architectures relatively quickly. (But cloud computing is a whole topic in its own right and, as such, shall be reserved for a future blog where I hope to also include some reflections on the recent tie-up with Akamai as I believe this could be crucial.)
Will all this help to sell more NetScalers? Almost certainly. I've not even talked about:
1) the fact that Citrix have developed configuration wizards to ease NetScaler deployments,
2) how easy it can be to convince an Access Gateway customer to upgrade to a NetScaler,
3) the imminent release of v9.0 where we can expect specific application integration and
4) that training programs are going online and will cost about 20% of what they used to.
There's a whole lot more to NetScaler than initially meets the eye, it's just a question of whether resellers (and customers) will see it.
Thursday, 23 October 2008
Mixed fortunes for CTXS in Q308
Citrix announced their Q3 2008 results today and, whilst revenue growth is impressive ($399 million compared to $350 million and therefore 14% up on Q3 last year), profits have taken a bit of a hit. It has not gone unnoticed that what previously was a seemingly endless stream of acquisitions has come to a complete stop lately. The last two announcements that I can think of involving 3rd party companies, Sepago and Marathon, have been, in Sepago's case, a technology purchase rather than a full-blown company purchase and, in Marathon's case, an OEM agreement.
This isn't meant as a criticism, quite the contrary. I feel Citrix have enough on their plate for the time being getting the Delivery Centre up and running properly (integration of NetScaler, live release of Workflow Studio, ironing out of XenDesktop imperfections etc.), so adding yet further technologies this year might be pushing the boat out slightly too far. Added to that, after the not inconsiderable investments over the past few years (NetScaler, XenSource et al) a period of consolidation, which I presume we are now in, was certainly to be expected.
Good to see, however, that both Citrix and VMWare are bucking the economic trends and posting pretty solid results, all considered. I certainly have an optimistic outlook for the rest of this year. Full details on Citrix's performance here.
This isn't meant as a criticism, quite the contrary. I feel Citrix have enough on their plate for the time being getting the Delivery Centre up and running properly (integration of NetScaler, live release of Workflow Studio, ironing out of XenDesktop imperfections etc.), so adding yet further technologies this year might be pushing the boat out slightly too far. Added to that, after the not inconsiderable investments over the past few years (NetScaler, XenSource et al) a period of consolidation, which I presume we are now in, was certainly to be expected.
Good to see, however, that both Citrix and VMWare are bucking the economic trends and posting pretty solid results, all considered. I certainly have an optimistic outlook for the rest of this year. Full details on Citrix's performance here.
Monday, 20 October 2008
Bad eggs
Last week I attended a successful event staged by one of our key customers, Nebulas Solutions, in London. They recently formed a division called Nebulas Virtualise, which focusses, as you might expect, on virtualisation and related technologies. Nebulas also employ what they call a "Technology Incubator" to help them decide which products should be incorporated into their corporate portfolio of "fully" approved offerings. Essentially, they thoroughly test a specific product themselves, then introduce it to their early adopter type customers and, if all goes well, they transfer it to the mainstream. RES Software, one of the vendors I am responsible for at COMPUTERLINKS, has just been included in this, alongside RTO Software, which many readers might already know of, plus, in the security space, Tufin.
The event was based mainly around delivering Virtual Desktops and was therefore headlined by VMWare. Citrix was also mentioned quite a bit throughout, but of course only in terms of their server-based computing solution and, rather unimaginatively, in the context of what Citrix XenApp CAN'T do. (I must admit it took some self-control not to point out to the VMWare speaker that Citrix have spent the last 2 years telling customers that, yes, of course XenApp is not the ideal solution for every scenario, hence the evolution of XenDesktop to solve some of those problems, but that it is without doubt the lowest cost option. I kept my trap shut...)
What did strike me, however, is that the VMWare pitch around VDI was exactly the same, and I mean almost a carbon copy, of what Citrix are preaching. I don't mean that is a bad thing, quite the opposite actually, with two companies pushing the same messaging, it means the market may grow that much quicker. Which solution a customer ultimately then goes with will be decided by a load of other factors but, if the concept of virtual desktops itself is a quantifiably sound investment, at least my personal future is safe(ish).
So VMWare and Citrix are saying the same thing and, as I heard at the above event, an analyst from Gartner predicted a couple of years back that every single new desktop will be virtualised by the end of 2010 (!), why are they currently the only two companies in the world capable of providing this? OK, OK, I'm sure there are one or two others such as the perennial Ericom, Quest or even Symantec (and probably Microsoft in the future) who will, perhaps justifiably, claim they can do it too, but in all honesty, VMWare and Citrix are the two vendors that immediately spring to most educated minds on the mention of virtual desktops.
So why on earth, and this is actually the main point of this article, are the Citrix and VMWare channels probably the least profitable, in terms of margin retention for distributors and resellers, of just about any software product there is?
Citrix and VMWare develop highly complex infrastructure-level software solutions yet it seems some of the channel are happy to sell licenses at about the same margin levels as they retain on hardware. Any Tom, Dick or Harry can assemble a PC these days, my brother-in-law made one for my parents last Christmas, yet when it comes to revolutionising the way a company actually accesses applications, and therefore, ultimately, how they do business, why are we happy to "knock out" the licenses as though they are PCs? And, added to that, there are only perhaps 3 or 4 companies who can do this! Compare that to the amount of PC makers there are.
The reasons, in my opinion, all boil down to the same thing - bad eggs. I suspect there are many fickle end users, quite happy to let one reseller do all the preparatory work (sometimes free of charge) and then just buy the licenses off the company who eventually submits the lowest bid, often one who has had absolutely nothing to do with the project (so no cost of sale) and probably never will. Secondly, the reseller community adopt the same approach with the distributors and, thirdly, the one or two distributors who let this happen.
In the current economic climate, perhaps some acceptance of pricing pressure is necessary, but it seems to me that there are certain organisations (and believe me, I really wish I could name them here but I'm sure I'd be sued if I did so please Do Not Sue me), who defy any sort of sensible, long-term thinking for the industry as a whole. They are quite happy to offer ridiculous pricing to both resellers and end users who do not deserve it, in the vain hope of meeting this month's revenue target, come hell or high water.
We work in a competitive industry, sure, and I suppose that, in a sense, it's every man for himself when the proverbial hits the fan, but I think we also have a responsibility to ourselves and to our chosen career paths and specific areas of technology, to ensure that some modicum of value remains. Citrix and VMWare are market-leading, "best-of-breed" solutions and are therefore expensive to buy - rightly so. In the same way as Mercedes-Benz and BMW cars are market-leadingly expensive.
Whilst the vendors carry some responsibility to protect their brand and prevent it becoming yet another commodity item, their hands are tied to a point - certainly from a legal point of view, so surely the buck stops with the channel. (On the other hand, don't the vendors have the option to pick and choose which organisations are accepted into their channels?) Please don't think I am advocating a market where price-fixing becomes acceptable, we need only look at British Airways to see the ramifications of that sort of behaviour, all I'm calling for is that the prices charged for those software solutions that a) have a high cost of sale in terms of training, staffing levels, up-front investment etc. and b) provide answers to problems only a handful of companies are skilled enough and experienced enough to do, are fair, appropriate to effort and at a level that ensures everyone's futures.
Currently, I'm afraid, I get the distinct feeling those few bad eggs are spoiling things for all of us, the market is degenerating into a smash n' grab fest and for those of us actually serious about these technologies, the desire to further invest is evaporating quickly.
The event was based mainly around delivering Virtual Desktops and was therefore headlined by VMWare. Citrix was also mentioned quite a bit throughout, but of course only in terms of their server-based computing solution and, rather unimaginatively, in the context of what Citrix XenApp CAN'T do. (I must admit it took some self-control not to point out to the VMWare speaker that Citrix have spent the last 2 years telling customers that, yes, of course XenApp is not the ideal solution for every scenario, hence the evolution of XenDesktop to solve some of those problems, but that it is without doubt the lowest cost option. I kept my trap shut...)
What did strike me, however, is that the VMWare pitch around VDI was exactly the same, and I mean almost a carbon copy, of what Citrix are preaching. I don't mean that is a bad thing, quite the opposite actually, with two companies pushing the same messaging, it means the market may grow that much quicker. Which solution a customer ultimately then goes with will be decided by a load of other factors but, if the concept of virtual desktops itself is a quantifiably sound investment, at least my personal future is safe(ish).
So VMWare and Citrix are saying the same thing and, as I heard at the above event, an analyst from Gartner predicted a couple of years back that every single new desktop will be virtualised by the end of 2010 (!), why are they currently the only two companies in the world capable of providing this? OK, OK, I'm sure there are one or two others such as the perennial Ericom, Quest or even Symantec (and probably Microsoft in the future) who will, perhaps justifiably, claim they can do it too, but in all honesty, VMWare and Citrix are the two vendors that immediately spring to most educated minds on the mention of virtual desktops.
So why on earth, and this is actually the main point of this article, are the Citrix and VMWare channels probably the least profitable, in terms of margin retention for distributors and resellers, of just about any software product there is?
Citrix and VMWare develop highly complex infrastructure-level software solutions yet it seems some of the channel are happy to sell licenses at about the same margin levels as they retain on hardware. Any Tom, Dick or Harry can assemble a PC these days, my brother-in-law made one for my parents last Christmas, yet when it comes to revolutionising the way a company actually accesses applications, and therefore, ultimately, how they do business, why are we happy to "knock out" the licenses as though they are PCs? And, added to that, there are only perhaps 3 or 4 companies who can do this! Compare that to the amount of PC makers there are.
The reasons, in my opinion, all boil down to the same thing - bad eggs. I suspect there are many fickle end users, quite happy to let one reseller do all the preparatory work (sometimes free of charge) and then just buy the licenses off the company who eventually submits the lowest bid, often one who has had absolutely nothing to do with the project (so no cost of sale) and probably never will. Secondly, the reseller community adopt the same approach with the distributors and, thirdly, the one or two distributors who let this happen.
In the current economic climate, perhaps some acceptance of pricing pressure is necessary, but it seems to me that there are certain organisations (and believe me, I really wish I could name them here but I'm sure I'd be sued if I did so please Do Not Sue me), who defy any sort of sensible, long-term thinking for the industry as a whole. They are quite happy to offer ridiculous pricing to both resellers and end users who do not deserve it, in the vain hope of meeting this month's revenue target, come hell or high water.
We work in a competitive industry, sure, and I suppose that, in a sense, it's every man for himself when the proverbial hits the fan, but I think we also have a responsibility to ourselves and to our chosen career paths and specific areas of technology, to ensure that some modicum of value remains. Citrix and VMWare are market-leading, "best-of-breed" solutions and are therefore expensive to buy - rightly so. In the same way as Mercedes-Benz and BMW cars are market-leadingly expensive.
Whilst the vendors carry some responsibility to protect their brand and prevent it becoming yet another commodity item, their hands are tied to a point - certainly from a legal point of view, so surely the buck stops with the channel. (On the other hand, don't the vendors have the option to pick and choose which organisations are accepted into their channels?) Please don't think I am advocating a market where price-fixing becomes acceptable, we need only look at British Airways to see the ramifications of that sort of behaviour, all I'm calling for is that the prices charged for those software solutions that a) have a high cost of sale in terms of training, staffing levels, up-front investment etc. and b) provide answers to problems only a handful of companies are skilled enough and experienced enough to do, are fair, appropriate to effort and at a level that ensures everyone's futures.
Currently, I'm afraid, I get the distinct feeling those few bad eggs are spoiling things for all of us, the market is degenerating into a smash n' grab fest and for those of us actually serious about these technologies, the desire to further invest is evaporating quickly.
Friday, 3 October 2008
Citrix lives! (Albeit in a cloud)
Returning, as I do, with a glistening wedding ring on my finger, I must now live up to my new status as a mature, upstanding, married member of the community. Losing my phone on honeymoon wasn't the best start but I'll put it down to teething problems.
When I (temporarily) signed off a couple of weeks ago, it was a matter of days before VMWorld and I was utterly convinced that Citrix would announce an acquisition of the leading HA and fault tolerance vendor Marathon. In the end it turned out not to be a full takeover but, almost as good, a tight OEM agreement, building Marathon's utterly brilliant technology (I saw this running a while back and was gobsmacked) into the synchronised release of XenServer v5.0.
It just goes to show that Citrix's claims of continued openness to partnership and upholding the DNA of industry co-operation that existed at XenSource before they were snapped up, do indeed ring very true. VMWare, as usual, have been developing their own HA solution, which will surely be hard-pushed to match the might of the XenServer/Marathon solution. HA with XenServer is now available in three flavours ("dialable" is the latest irritating IT buzzword for this):
1) automatic re-start of VMs (which VMWare have had for some time),
2) component level fault tolerance (e.g. a disk or network card goes down and Marathon moves everything across instantaneously to the back-up server with little to no impact), effective within a radius of about 50km - depending on the speed of the connection obviously,
3) global full system "five nines" availability providing as good as zero downtime (Marathon has a customer who have been running EverRun for 9 years and they have had just 11 seconds downtime in that period). This is available now for physical servers, Q1 09 in a virtual environment.
So my prognosis below that Citrix would upstage VMWare yet again wasn't quite borne out to the extent I had hoped but I still think there was almost as much coverage from VMWorld about them as there was VMWare.
Which leads me to another topic that hit the headlines whilst I was away. Apparently Citrix was to be bought by Microsoft. There have been several references to this, e.g. here, here and here, the latter of which led to two industry heavyweights, Brian Madden and Doug Brown, rather amusingly having a bit of a tiff about it, but, as we now know, it has all fizzled out to nothing. Again.
Microsoft buying Citrix must be one of the oldest rumours in the industry and it never seems to go away. Surely we must only ask ourselves the question why? What on earth would Microsoft get from buying Citrix? Sure, Citrix have better technology in most areas but would MS's shareholders really agree to a CapEx of what would probably amount to around 5-6 billion dollars for a slight improvement on what they already have? I doubt it, even if they do have pots of cash to spend.
And others? Brian even suggested VMWare as a potential suitor (and got completely lambasted for it in the responses - entirely unfairly in my opinion). Other names I've heard thrown into the hat are Symantec, IBM, Cisco, Google, Oracle and just about anyone else that could feasibly afford them.
Honestly? I think Citrix's stock is well down at the moment (just over $22 at the time of writing), as are many companies', so it would certainly be a reasonably good time to launch even a hostile takeover, however it's not the first time it's been down in the 20's and nothing happened the last few times. Perhaps the difference now is that, where Citrix were a delectable little canape a few years ago, they are rapidly becoming quite a mouthful these days. Their relentless attempts to establish themselves as an infrastructure player (alongside some of those names mentioned above) and an industry standard in application delivery and virtualisation has been quite a ride for us distributors. For this reason, I don't think they are anywhere near the finished article yet and I can't see a glaringly obvious candidate out there at the moment for whom they would provide a fully rounded, market-beating solution, even in an "embellishment" role. Not yet anyway.
Apparently, the next step on this journey is the "Cloud" - otherwise known as the Internet. (How on earth did we manage to re-brand the humble T'interweb and turn it into a new market area by the way?) Both VMWare and Citrix announced some cloudy stuff at VMWorld and I need to get my head around it before I blog with any authority on it, but I will say this. Mention applications and data in the same sentence as the word Internet to most security people and they will shudder. Never heard of TK Maxx, Marks & Spencer, British Military, [enter your favourite household name]?
When I (temporarily) signed off a couple of weeks ago, it was a matter of days before VMWorld and I was utterly convinced that Citrix would announce an acquisition of the leading HA and fault tolerance vendor Marathon. In the end it turned out not to be a full takeover but, almost as good, a tight OEM agreement, building Marathon's utterly brilliant technology (I saw this running a while back and was gobsmacked) into the synchronised release of XenServer v5.0.
It just goes to show that Citrix's claims of continued openness to partnership and upholding the DNA of industry co-operation that existed at XenSource before they were snapped up, do indeed ring very true. VMWare, as usual, have been developing their own HA solution, which will surely be hard-pushed to match the might of the XenServer/Marathon solution. HA with XenServer is now available in three flavours ("dialable" is the latest irritating IT buzzword for this):
1) automatic re-start of VMs (which VMWare have had for some time),
2) component level fault tolerance (e.g. a disk or network card goes down and Marathon moves everything across instantaneously to the back-up server with little to no impact), effective within a radius of about 50km - depending on the speed of the connection obviously,
3) global full system "five nines" availability providing as good as zero downtime (Marathon has a customer who have been running EverRun for 9 years and they have had just 11 seconds downtime in that period). This is available now for physical servers, Q1 09 in a virtual environment.
So my prognosis below that Citrix would upstage VMWare yet again wasn't quite borne out to the extent I had hoped but I still think there was almost as much coverage from VMWorld about them as there was VMWare.
Which leads me to another topic that hit the headlines whilst I was away. Apparently Citrix was to be bought by Microsoft. There have been several references to this, e.g. here, here and here, the latter of which led to two industry heavyweights, Brian Madden and Doug Brown, rather amusingly having a bit of a tiff about it, but, as we now know, it has all fizzled out to nothing. Again.
Microsoft buying Citrix must be one of the oldest rumours in the industry and it never seems to go away. Surely we must only ask ourselves the question why? What on earth would Microsoft get from buying Citrix? Sure, Citrix have better technology in most areas but would MS's shareholders really agree to a CapEx of what would probably amount to around 5-6 billion dollars for a slight improvement on what they already have? I doubt it, even if they do have pots of cash to spend.
And others? Brian even suggested VMWare as a potential suitor (and got completely lambasted for it in the responses - entirely unfairly in my opinion). Other names I've heard thrown into the hat are Symantec, IBM, Cisco, Google, Oracle and just about anyone else that could feasibly afford them.
Honestly? I think Citrix's stock is well down at the moment (just over $22 at the time of writing), as are many companies', so it would certainly be a reasonably good time to launch even a hostile takeover, however it's not the first time it's been down in the 20's and nothing happened the last few times. Perhaps the difference now is that, where Citrix were a delectable little canape a few years ago, they are rapidly becoming quite a mouthful these days. Their relentless attempts to establish themselves as an infrastructure player (alongside some of those names mentioned above) and an industry standard in application delivery and virtualisation has been quite a ride for us distributors. For this reason, I don't think they are anywhere near the finished article yet and I can't see a glaringly obvious candidate out there at the moment for whom they would provide a fully rounded, market-beating solution, even in an "embellishment" role. Not yet anyway.
Apparently, the next step on this journey is the "Cloud" - otherwise known as the Internet. (How on earth did we manage to re-brand the humble T'interweb and turn it into a new market area by the way?) Both VMWare and Citrix announced some cloudy stuff at VMWorld and I need to get my head around it before I blog with any authority on it, but I will say this. Mention applications and data in the same sentence as the word Internet to most security people and they will shudder. Never heard of TK Maxx, Marks & Spencer, British Military, [enter your favourite household name]?
Thursday, 11 September 2008
On a somewhat more personal note...
There comes a time in most men's lives when they must kiss a fond farewell to youth, freedom and singleton-ism (if that's even a word) and devote their undivided attention to the future mother of their children. That time has now, happily, arrived for the otherwise completely devoted creator of this blog. It is with perhaps a touch of excited nervosity, but yet every confidence, that, at about 1:00pm on Friday 19th September 2008, I will be standing by the altar of the Catholic church in Newmarket, about to turn round and watch my esteemed marketing colleague at COMPUTERLINKS, Maria Curley, walking up the aisle, preparing to admit she wants to spend the rest of her life with me.
I honestly couldn't be a happier chappy but, ultimately, something must give. Blogs and new wives just don't mix, so the blog will just have to get used to life with its master as a married man and take a break for a couple of weeks.
I shall resume my ramblings upon my return when, I happen to know, there will most certainly be one or two things to discuss. (It's VMWorld next week, don't forget, and I can safely say now that Citrix may well out-manoeuvre VMWare on the announcement front once again - the first time being the XenSource acquisition the day before VMWare's IPO. Sorry, but I am not permitted say any more than that at this stage.)
Until then, I bid thee goodbye. I promise to think of you all, suffering the early UK winter, whilst I contemplate whether to take a 6 or a 7 iron on the first par 3 of the Old Course at Vilamoura.
My tips for a wife? Find someone who organises events for a living and has previously been called The Incredible Sleeping Woman whilst on holiday. You need do next to nothing for the wedding and you get to play golf on your honeymoon while she sleeps. That, my friends, is the perfect wife!
Adios amigos.
Green IT not a spin-led "greenwash"
Again, just like the "Virtualisation - a bit of a snapshot" article below, this piece was written a while ago for a broader audience than this blog typically addresses, but I would like to post it nonetheless.
Whilst doubters remain, cynically citing vendor marketing spin, concerning the real impact that such things as virtualisation have on energy consumption and carbon footprints, companies such as Citrix do in fact provide undeniable and measurable improvements.
Getting fewer servers to do the same job, using XenServer for example, is the fundament. Other products though, particularly within the Citrix portfolio, such as XenApp Platinum or Citrix Online, enable secure remote access and fuel-saving work-from-home environments and these also contribute substantially.
As most IT managers are now aware, server virtualisation vastly reduces the amount of servers required to service the environment. This impacts not only energy consumption in powering the actual devices, but also the considerable air conditioning required to keep the data-centres at optimum temperature.
Citrix incorporate a product called Provisioning Server into the Platinum version of their XenServer solution. Taking things to the extreme, this could provide another stepping stone towards the Holy Grail of using energy only when that energy is actually required. Instead of having a fully functional data-centre running all day and all night, regardless of whether it's actually in use or not, Provisioning Server, in tandem with one or two other solutions, provides the capability to fire up servers and take them down again according to time of day and the usage requirements of the organisation.
In an ideal world, an administrator could effectively dismantle the data-centre, or reduce it down to a bare minimum, at 10pm and then start it all back up again at 6am the next morning. This could involve the servers running the same workloads as the previous day or indeed completely different ones. Provisioning Server transforms a server from an SAP server, say, into a Navision server, in just a few minutes. The workloads are stored on virtual disks at the back-end and streamed out to the hardware as and when required.
Taking this one step further, Citrix EdgeSight provides metrics around the end user experience and another complementary vendor, RES Software, manages hardware remotely. So, rather than performing these tasks manually, the administrator could set up rules according to the metrics being delivered back from the EdgeSight Server and then configure RES Wisdom to shut down the servers in the evening when demand goes below a certain point. The next morning, when demand begins to grow again, RES wakes up the machines and Provisioning Server churns out the workloads. Again, demand could also determine what type of server each turns out to be.
These products will undoubtedly integrate closely with one another to enable something approaching a 100% dynamic, on-demand data-centre. Citrix’s soon-to-be-released Workflow Studio will then add a further piece to the pie, providing a simple graphical interface for these processes to be simply and quickly put into place. Throw in remote infrastructure management solutions such as Avocent and you create a remarkably flexible, energy-efficient data-centre.
Moving to the desktop, when you consider a PC uses around 75 watts of power and a WYSE thin client requires about 10 or 15% of that, CFOs will surely demand increasingly sound justification from their CTOs for not choosing the “server-based computing” or application delivery methods over traditional user scenarios.
More dedicated device management also comes into the frame – not only in the data-centre but also on the desktop. RES Software, for example, can enforce a power-down of PCs left on overnight. For a company with 150 employees leaving just 15% of the PCs running overnight, the total cost of the software is offset by the energy savings in just 9 months. Such is the effect of high energy prices, the software pays for itself in under a year on that one functionality alone.
IT departments are now becoming cost centres with their own profit & loss accounts, making energy reduction IT’s problem. Reed Managed Services are Citrix’s flagship case study for green IT and energy usage reduction. Admittedly other non IT-related measures were also implemented, but with 64-bit Citrix virtualisation solutions (and rolling out thin clients to almost all their users), Reed: a) got 100 users on a blade instead of 29, b) made their company 100% carbon neutral, c) saved 26% of their IT budget and d) won them a Green Oscar.
Not bad for vendor marketing spin!
Whilst doubters remain, cynically citing vendor marketing spin, concerning the real impact that such things as virtualisation have on energy consumption and carbon footprints, companies such as Citrix do in fact provide undeniable and measurable improvements.
Getting fewer servers to do the same job, using XenServer for example, is the fundament. Other products though, particularly within the Citrix portfolio, such as XenApp Platinum or Citrix Online, enable secure remote access and fuel-saving work-from-home environments and these also contribute substantially.
As most IT managers are now aware, server virtualisation vastly reduces the amount of servers required to service the environment. This impacts not only energy consumption in powering the actual devices, but also the considerable air conditioning required to keep the data-centres at optimum temperature.
Citrix incorporate a product called Provisioning Server into the Platinum version of their XenServer solution. Taking things to the extreme, this could provide another stepping stone towards the Holy Grail of using energy only when that energy is actually required. Instead of having a fully functional data-centre running all day and all night, regardless of whether it's actually in use or not, Provisioning Server, in tandem with one or two other solutions, provides the capability to fire up servers and take them down again according to time of day and the usage requirements of the organisation.
In an ideal world, an administrator could effectively dismantle the data-centre, or reduce it down to a bare minimum, at 10pm and then start it all back up again at 6am the next morning. This could involve the servers running the same workloads as the previous day or indeed completely different ones. Provisioning Server transforms a server from an SAP server, say, into a Navision server, in just a few minutes. The workloads are stored on virtual disks at the back-end and streamed out to the hardware as and when required.
Taking this one step further, Citrix EdgeSight provides metrics around the end user experience and another complementary vendor, RES Software, manages hardware remotely. So, rather than performing these tasks manually, the administrator could set up rules according to the metrics being delivered back from the EdgeSight Server and then configure RES Wisdom to shut down the servers in the evening when demand goes below a certain point. The next morning, when demand begins to grow again, RES wakes up the machines and Provisioning Server churns out the workloads. Again, demand could also determine what type of server each turns out to be.
These products will undoubtedly integrate closely with one another to enable something approaching a 100% dynamic, on-demand data-centre. Citrix’s soon-to-be-released Workflow Studio will then add a further piece to the pie, providing a simple graphical interface for these processes to be simply and quickly put into place. Throw in remote infrastructure management solutions such as Avocent and you create a remarkably flexible, energy-efficient data-centre.
Moving to the desktop, when you consider a PC uses around 75 watts of power and a WYSE thin client requires about 10 or 15% of that, CFOs will surely demand increasingly sound justification from their CTOs for not choosing the “server-based computing” or application delivery methods over traditional user scenarios.
More dedicated device management also comes into the frame – not only in the data-centre but also on the desktop. RES Software, for example, can enforce a power-down of PCs left on overnight. For a company with 150 employees leaving just 15% of the PCs running overnight, the total cost of the software is offset by the energy savings in just 9 months. Such is the effect of high energy prices, the software pays for itself in under a year on that one functionality alone.
IT departments are now becoming cost centres with their own profit & loss accounts, making energy reduction IT’s problem. Reed Managed Services are Citrix’s flagship case study for green IT and energy usage reduction. Admittedly other non IT-related measures were also implemented, but with 64-bit Citrix virtualisation solutions (and rolling out thin clients to almost all their users), Reed: a) got 100 users on a blade instead of 29, b) made their company 100% carbon neutral, c) saved 26% of their IT budget and d) won them a Green Oscar.
Not bad for vendor marketing spin!
Subscribe to:
Posts (Atom)